Hardware that makes AI sovereign.

Dell Precision Tower

Industrial shape and dense performance: the Dell Precision Tower is a proven base for OPAIRS deployments in production environments.

Precision T7920 → T7960 · multiple generations in use

Dell — performance for industrial AI

• •Dual Xeon and Xeon-W configurations for parallel LLM inference and data processing
• •Scalable RAM and storage options from entry setup to full OPAIRS data lake expansion
• •Enterprise support and extended service levels for production operations without cloud fallback

Dell Precision Tower

Service-friendly interior for modular GPU upgrades, important when phase 2 adds specialized inference cards or fine-tunes local models on dedicated hardware.

Precision T7920 → T7960 · multiple generations in use

Dell — security & manageability

• •GPU-accelerated workloads for OPAIRS BI dashboards and real-time anomaly detection on the shop floor
• •High reliability in 24/7 operation, critical for air-gapped installations without external support fallback
• •Compatible with OPAIRS security-by-design architecture aligned with ISO 27001 / 27701 and the EU AI Act

HP Z Workstation

The HP Z series is a proven platform in the OPAIRS ecosystem. First prototypes started on HP Z8G4; current deployments use HP Z8 Fury G5.

Z8 G4 (Prototyp) → Z8 Fury G5 (aktuelle Basis)

HP — proven platform since prototype phase

• •The HP Z8G4 is one of the two original prototype servers used for first OPAIRS LLM inference and IOMETE data lake tests
• •Flexible PCIe expansion for dedicated AI accelerators, foundation for the OPAIRS high-performance product line from 2027
• •Manageable platform directly compatible with OPAIRS NIS2 and EU AI Act compliance layer

HP Z8 Fury

Thoughtful internal layout with fast component access; during OPAIRS rollout, systems are customized on site. Z8 Fury G5 is planned as a base for the 2027 high-performance line.

Z8 G4 (Prototyp) → Z8 Fury G5 (aktuelle Basis)

HP — scaling without vendor lock-in

• •Efficient cooling architecture for stable AI inference in 24/7 production
• •Long-term partner perspective with HP at sufficient OPAIRS scale for strategic supply chain stability
• •Complemented by certified refurbished options for cost-sensitive pilot customers in the DACH region

Lenovo ThinkStation

Reduced design with strong compute reserves: the second original OPAIRS prototype was based on Lenovo P920; today ThinkStation PX bridges mid and high performance.

P720 / P920 (Prototyp) → ThinkStation PX (aktuell)

Lenovo — second pillar in the OPAIRS ecosystem

• •Lenovo P920 is one of the first two OPAIRS development servers, proven for data processing and LLM inference under real conditions
• •Configurable for industry-specific requirements in BIM, simulation and industrial computer vision
• •High operational stability for critical production phases without external SaaS dependency

Lenovo ThinkStation PX

Tool-free component access, ideal for on-site OPAIRS installations where technicians commission and extend systems directly at customer premises.

P720 / P920 (Prototyp) → ThinkStation PX (aktuell)

Lenovo — from pilot to team rollout

• •Scalable from a single workstation to multi-stage OPAIRS rollout across multiple production sites
• •Suitable for hybrid on-premise environments with central OPAIRS orchestration and distributed edge nodes
• •Stable for long LLM inference and data pipeline processes, critical for OPAIRS data lake operations in production